Text this: Global Multi-Phase Path Planning Through High-Level Reinforcement Learning